Alcohol Guidelines Review Report from the Guidelines development group to the UK Chief Medical Officers

Transcription

1 Alcohol Guidelines Review Report from the Guidelines development group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) January 2016

2 2 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) Executive Summary Health evidence expert group and behavioural expert group, In late 2012 the UK Chief Medical Officers commissioned two expert groups to consider whether previous alcohol guidelines should be updated. Both expert groups worked through 2013 and reported to the UK CMOs in February The Health evidence expert group examined the evidence 1 from 44 systematic reviews and meta-analyses published since the 1995 Sensible drinking report 2, and consulted experts recently involved in the updating of the Australian and Canadian alcohol guidelines. 3. The group concluded 3 that there is significant new, good quality evidence available on the effects of alcohol consumption on health, which was not available at the time of the 1995 review. This applies for both men and women. In particular, stronger evidence has emerged that the risk of a range of cancers, especially breast cancer, increases directly in line with consumption of any amount of alcohol. 4. Meta-analyses have identified that for some conditions, notably ischaemic heart disease (IHD), drinking alcohol at low levels may have a protective effect (compared to not drinking), particularly for all-cause mortality. However, the group noted that: any potential protective effect seems mainly relevant to older age groups; unresolved confounding and health selection (for instance, the health of people who can afford to drink more in older age may be better than those who do not) may explain a substantial part of the protection observed; mortality from IHD is continuing to decrease substantially; and the peak of any protective effect is achieved at very low levels of consumption (around one unit a day). 5. The group therefore concluded that the evidence supporting protective effects today is now weaker than it was at the time of the 1995 report and that there are substantial uncertainties around direct attribution to alcohol of the level of protection still observed. Taking this into account alongside all the known acute and chronic risks to health from drinking even at low levels, supports the conclusion of the group that there is no justification for recommending drinking on health grounds, nor for starting drinking for health reasons. 6. For alcohol consumption in pregnancy, more recent systematic review evidence concludes that the risks of low birth weight, preterm birth, and being small for gestational age all may increase above consumption of 1-2 units/day. Overall consideration of the evidence published since the NICE review in 2008 supports precautionary guidance that it is safest to avoid drinking in pregnancy. 1 CMO Alcohol Guidelines Review A summary of the evidence of the health and social impacts of alcohol consumption; CMO Alcohol Guidelines Review Mapping systematic review level evidence; both reviews were written by the Centre for Public Health, Liverpool John Moores University 2 Department of Health (1995) Sensible drinking: Report of an inter-departmental working group 3 CMO s Review of Alcohol Guidelines: Conclusions of the Health Evidence Expert Group

3 Executive Summary 3 7. For alcohol consumption by young people under 18, the group found very little new evidence had been published since guidance was published by the English CMO in 2009 and subsequently by the other countries CMOs. 8. The Behavioural expert group found little evidence 4 regarding the impact of any guidelines in changing health behaviours. It concluded 5 that there are, however, some general principles both from evidence reviews in the field of public health and from other fields (social marketing, goal setting, behavioural psychology etc.) that could be used to maximise understanding and acceptance, if the UK does introduce new guidelines based on new evidence. Based on these, the group recommended that: the justification for any guidelines should be clearly spelt out and they should be: simple, specific, measurable, timebound; and realistic 9. Neither of the expert groups felt there was adequate justification for having separate guidelines for different age or social groups. Guidelines development group, In early 2014, the CMOs asked members from the two previous groups to combine, to advise on the most appropriate methodological approach to developing guidelines (in particular comparing the Australian and Canadian approaches used to develop their recent guidelines), and to advise on appropriate guidelines. 11. Terms of reference for the Guidelines development group are at Annex A. 12. The group considered carefully its approach and the evidence for advice to the public on the health risks from drinking. It based its advice on the following principles: People have a right to accurate information and clear advice about alcohol and its health risks, and There is a responsibility on Government to ensure this information is provided for citizens in an open way, so that they can make informed choices. 13. As the Guidelines development group, we feel that throughout the dissemination of the guidelines it is critical this is seen as a process to inform individuals so that they can take informed decisions about their own drinking. 14. Australia used an absolute risk approach whereby the guidelines were set at a point above which individuals would experience a 1% lifetime risk of dying due to alcohol. 6 Canada used a relative risk approach which set the guidelines at a point at which alcohol harms and benefits were balanced, in terms of the numbers of deaths caused and prevented. 15. The Guidelines development group requested Public Health England to put out a public call for tenders to: 4 CMO Alcohol Guidelines Review A summary of the evidence on understanding and response to public health guidelines, Centre for Public Health, Liverpool John Moores University 5 CMO Alcohol Guidelines Review: Report from the Behavioural Expert Group 6 Currently the average lifetime risk of dying from an alcohol-related cause in the UK is about 4%

4 4 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) i. provide quantified risk estimates for the mortality and morbidity (the latter defined as person-specific hospital admissions) associated with different levels and patterns of alcohol consumption for drinkers in the UK; and ii. report on the guideline thresholds that would be derived from applying approaches similar to those used in Canada and Australia to UK mortality and morbidity risk estimates. 16. Sheffield University was awarded the tender for this work, and used the Sheffield Alcohol Policy Model as the basis for its report The Guidelines development group also consulted additional experts (see list of contributors at Annex B); and were briefed about recent findings from focus group research, which explored public understanding of the current guidelines. Research with the public commissioned by Public Health England explored their response to the draft guidelines below. 18. The Guidelines development group recommends that women who are pregnant or planning a pregnancy should be advised that the safest approach is not to drink alcohol at all. There is no scientific basis for setting a limit below which alcohol consumption will not harm the fetus. There is evidence that the second part of the current English guidelines (If they (pregnant women) do choose to drink, to minimise the risk to the baby, they should not drink more than one to two units of alcohol once or twice a week and should not get drunk) can be seen as either having equal weight with or being inconsistent with the first part (Pregnant women or women trying to conceive should avoid drinking alcohol), and there is no clear scientific evidence to support a quantified limit for drinking in pregnancy such as that in the second part of the current advice. 19. This guidance is precautionary. It will be important to inform women that the risk of harm to the baby is likely to be low if they have drunk only small amounts of alcohol before being aware of a pregnancy. Nevertheless, we cannot rule out the risks altogether. 20. Harmonising the pregnancy guidelines across the UK, and making them consistent with those of other jurisdictions, would be helpful in delivering a clear and scientifically credible message. 21. The Guidelines development group recommends a new weekly guideline for regular drinking, and, in addition, new guidance in the form of narrative advice on reducing the shortterm risks from single occasion drinking. The guideline for weekly consumption is designed to reduce both the risk of deaths from regular drinking and the chronic, long term health harms that alcohol causes. This recommendation for a low risk level is the same for men and women. The new advice concerning single occasion drinking is designed to enable individuals to reduce their risks of acute, short-term harms from drinking. 22. There is evidence that there is currently some confusion about weekly and daily guidelines. Health Survey for England data (2011) suggest that most people who drink do so infrequently, with the majority of the population drinking on no more than 2 days per week. The current guidelines of regularly no more than 3 to 4 units per day for men and 2 to 3 for women can be seen as inappropriate (and therefore largely ignored) by those who only drink, for example on Fridays and Saturdays or on special occasions. The formulation of 3 to 4 and 2 to 3 can also be seen as confusing and potentially suggesting an optimum range rather than a limit for regular drinking. 23. The Sheffield model, which included a number of sensitivity analyses and estimates based on numbers of days drinking a week, shows that roughly similar conclusions can 7 Mortality and morbidity risks from alcohol consumption in the UK: Analyses using the Sheffield Alcohol Policy Model (v.2.7) to inform the UK Chief Medical Officers review of the UK lower risk drinking guidelines

5 Executive Summary 5 be reached from using the Canadian relative risk approach and an absolute risk approach adapted from that used in Australia. 24. As well as taking account of the point at which the risk of death from alcohol outweighs any potential health benefits, the group s report also took account of research showing levels of drinking where alcohol would be expected to cause an overall 8 1% lifetime risk of death in setting the proposed guideline on regular drinking. The guideline is at a level around or a little below this 1% risk. 25. Hence, the Guidelines development group recommend that for men and women who drink regularly or frequently i.e. most weeks, the guideline should be that you are safest not to drink regularly more than 14 units of alcohol per week. 26. The group s view is that, the weekly guideline on regular drinking also requires an additional recommendation, concerning the need to avoid harmful regular heavy drinking episodes, as there is clear evidence that such a pattern of drinking increases risk to health. Hence, we also recommend that it is best, if you do drink as much as 14 units per week, to spread this evenly over 3 days or more. If you have one or two heavy drinking sessions, you increase your risks of death from long term illnesses and from accidents and injuries. 27. In the group s judgement, there are strong arguments on scientific grounds related to the extensive health harms caused by alcohol (discussed more fully in the report) why a guideline for a low risk level of drinking in the UK should be recommended not to be higher than 14 units weekly for both women and men. The advice takes account of a likely previous overestimation and substantial uncertainties in the current estimate of protective effects directly attributable to drinking, as well as the greater evidence available on risks of heart disease, liver disease and certain cancers. The specific selection of 14 units is based on: careful consideration of the results of the scientific modelling, the detailed analysis of risks to the UK population from drinking, consideration of what is likely to be acceptable to be considered low risk, and the need to have a clear, understandable and specific message. 28. The recommendation not to exceed 14 units is both pragmatic and evidence-based. 29. The Guidelines development group recommends that the CMOs should give advice on short term health risks of alcohol related to any single drinking occasion: there are short term health risks, such as accidents and injuries, (and death caused by accidents or injuries) for anyone drinking to levels that cause intoxication. these risks can arise for people drinking within the weekly guidelines for regular drinking, if they drink too much or too quickly on a single occasion and for people who drink at higher levels, whether regularly or infrequently. the advice should cover the nature of the risks relating to accident and injury and actions people can take to reduce the risks of injury and accident, in addition to limiting their drinking. 8 Unlike the absolute risk method used in Australia, the method used in the Sheffield model does take account of deaths prevented, e.g. from ischaemic heart disease. So the 1% lifetime risk of death is a net risk for the population taking account of both deaths incurred and prevented.

6 6 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) 30. The group does not recommend providing an additional guideline level for low risk drinking on any one occasion in recognition of the individual variation of short term risks and the ability to reduce risks by other means. 31. The modelling shows that although the risks of alcohol consumption vary between men and women, between different types of consumption and harm, and between different age groups, there is good evidence to justify recommending a single low risk threshold for men and women. But there is insufficient consistent evidence relating to age to be able to recommend different low risk guidelines for different age groups. 32. In providing further information and narrative about the risks of alcohol, the CMOs could usefully provide a general explanation about how the risks might differ between individuals; for example because of effects of age, body size, familiarity with alcohol, the risks of breast cancer for women, or the risks of falls among older people. 33. The Guidelines development group has built on the work of the previous two expert groups. The group agreed that the weight of evidence to date suggests that: The benefits for heart health of drinking alcohol are less, and apply to a smaller group of the population, than previously thought. The Sheffield report commissioned for the expert group included a UK analysis which has found that the net protective effect 9 from mortality that may be attributable to drinking regularly at low levels appears now to be significant only for women aged 55+ (with men aged over 55+ showing such a net protective effect only of negligible size). The fact that the adverse effects of drinking alcohol on the risk of a range of cancers has now been strongly established, has the important consequence of the need to communicate that drinking any amount of alcohol regularly, does increase the risks of such serious harms (even though the risks of cancer attributable to alcohol are normally low if drinking is within the proposed guideline levels). However there is now evidence of an increased risk of certain cancers even at low levels of consumption. Both the above points taken together mean that we recommend to the CMOs that there is a need to be clear as a core message in future communications that the new guidelines are for low risk drinking not safe drinking. And that the vast majority of the population can reduce health risks further if they reduce drinking below the guideline levels, or do not drink at all. 34. The approach taken by the Guidelines development group has been to propose guidelines and advice to inform the public about the known health risks of different levels and patterns of drinking, particularly for people who want to know how to keep long term health risks from regular drinking of alcohol low. The regular drinking guideline provides advice which most of the population can use to keep their long term health risks low. 35. Many people, for personal or cultural reasons, drink no alcohol at all, most drink very moderately and only once or twice a week or less; and many drink for positive social and other benefits they perceive from drinking, including some who drink above low risk guideline levels. People vary in how they metabolise or react to alcohol, so people can be affected differently by drinking similar amounts. 36. Individuals will make their own judgements as to risks they are willing to accept from alcohol, whether to drink alcohol, and how much and how often to drink. To help people make informed judgements, we have provided a narrative with as much information as possible about the basis of the guidelines and the advice. 9 Taking account of mortality from all causes. For other groups, risk reductions for death from cardiovascular conditions are either largely or wholly outweighed by risk increases for other conditions.

7 Executive Summary The Guidelines development group recommend that as well as publishing the two specific new guidelines on pregnancy and regular drinking and the narrative advice on reducing the short-term risks from single drinking occasions, the CMOs publish a more extensive narrative about the basis for the new guidelines, and communicate clearly that: the risk of a number of cancers increases from any level of regular drinking; there is good evidence that cardio-protective effects have previously been overestimated and there are substantial uncertainties around the level of protection still observed; and the net cardio-protective effects from mortality attributable to drinking regularly at low levels are likely to be limited in the UK to women over the age of There is some limited evidence at present of the effects of having days free from alcohol consumption routinely for all drinkers. However, within a pattern of regular consumption, we recommend that adopting alcohol free days may be a useful way for drinkers to moderate their consumption. 39. Evidence is clear that there are other situations where it is advisable not to drink alcohol at all, or to exercise special caution, such as before and during driving, before, during, or directly after physical sport, before using machinery, before working where functioning in work would be adversely affected by alcohol, or when taking medication for which alcohol is contraindicated. 40. We recommend that the Government should run supportive social marketing campaigns for the public, developed in collaboration with the public. There should be a well funded Big Launch campaign. Campaigns should include education about both short and long term risks and alcohol units. 41. We recommend that the Department of Health works with health professionals and experts to review its guidance on higher risk drinking levels, in light of the new evidence underlying this report. 42. We recommend that health warnings and consistent messaging appear on all alcohol advertising, products, and sponsorship. 43. We recommend that systematic research be commissioned into the understanding, acceptability and uses of the new guidelines by the public, health professions and alcohol industry, including the impact of the supportive social marketing campaigns we recommend.

8 8 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) Alcohol Guidelines Review Report from the Guidelines development group to the UK Chief Medical Officers Final Report CONTENTS Page Introduction 9 Regular drinking advice 9 What evidence has changed on regular drinking guidelines? 12 Weekly or daily guidelines? 13 The basis for determining a low risk threshold for regular drinking 13 Is there a level of safe drinking? Are there acceptable risks from 14 alcohol? Individual variation 16 Men and women 17 Reduced risks from low levels of alcohol consumption 18 Harmful effects of alcohol and consideration of risks at different levels of 19 consumption Single occasions of drinking 23 Drinking and pregnancy 27 Other general considerations on the proposed guidelines and related advice 30 Situations where it is advisable not to drink alcohol at all 30 How clinicians might use new guidelines in public health advice 30 How might guidelines support changes at the community 33 and population level, e.g. through social marketing campaigns? The possible use of new technology to help people understand risks 33 and keep track of their own drinking Should we continue to use the current UK unit of alcohol? 34 Annexes Annex A Guidelines development group terms of reference 36 Annex B Members of the Guidelines Development Group and 38 additional experts consulted Annex C Effectiveness of, and engagement with, digital alcohol 39 interventions to reduce heavy drinking

9 Final Report 9 Alcohol Guidelines Review Report from the Guidelines development group to the UK Chief Medical Officers Final Report Introduction Rationale underpinning the report the overall approach to the draft guidelines 1. Previous reports from the Health evidence expert group and the Behavioural expert group set out the evidence behind their recommendation to develop new alcohol guidelines. 2. The Guidelines development group considered carefully its approach in reviewing the current guidelines and the evidence for advice to the public on the health risks from drinking. In proposing up-to-date guidance for the UK public, it based its advice on the following principles: People have a right to accurate information and clear advice about alcohol and its health risks There is a responsibility on Government to ensure this information is provided for citizens in an open way, so they can make informed choices Evidence on public understanding and response to guidelines 3. While there is little systematic evidence about the effectiveness of guidelines in changing health behaviours, there are some general lessons both from evidence reviews and from other fields (social marketing, goal setting, behavioural psychology etc.) which should be taken into account in introducing new guidelines based on new health evidence: the justification for any guidelines should be clearly spelt out and they should be: simple, specific, measurable, timebound; and realistic 4. It is crucial to present important public health information in a usable and comprehensible way that reflects the evidence base on health risk and the evidence base on effective communication of health messages

10 10 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) Regular drinking advice [this applies for people who drink regularly or frequently i.e. most weeks] New weekly guideline The Chief Medical Officers guideline for both men and women is that: You are safest not to drink regularly more than 14 units per week, to keep health risks from drinking alcohol to a low level If you do drink as much as 14 units per week, it is best to spread this evenly over 3 days or more. If you have one or two heavy drinking sessions, you increase your risks of death from long term illnesses and from accidents and injuries. The risk of developing a range of illnesses (including, for example, cancers of the mouth, throat and breast) increases with any amount you drink on a regular basis If you wish to cut down the amount you re drinking, a good way to help achieve this is to have several drink-free days each week. Evidence underpinning the advice 5. There is up-to-date epidemiological evidence on risks of health harm and mortality that were analysed for the UK in a report commissioned by Public Health England from the Sheffield Alcohol Research Group, University of Sheffield. 6. Some key findings are shown below based on both levels and frequency of drinking. 10 Tables 1 and 13 are taken from the Sheffield report. 10 The Sheffield report describes the Canadian and Australian methods: (1) In Canada, the guideline was based around epidemiological evidence suggesting that low levels of alcohol consumption are associated with reduced annual risk of mortality when compared with not drinking (i.e. there is some evidence that low levels of alcohol consumption provide a protective effect ). The Canadian guidelines for average daily consumption were thus set at the level at which risks of drinking were equivalent to those of abstaining from alcohol. In other words, the threshold level was chosen such that, at the population level, the estimated harmful effects and the estimated protective effects were counterbalanced equally against each other and net mortality risk was the same as if everyone abstained from alcohol. (2) In Australia, an alternative approach was used which focused on the absolute mortality risk due to drinking compared to the mortality risks from other causes. Thus the Australian guideline was set such that if the population all drank at that level, 1% of annual deaths would be attributable to alcohol. Selection of this 1% level was informed by guidance and regulations relating to other environmental and health risks and also by risks which appear to be acceptable to the public for other activities (e.g. the risk associated with driving a car regularly).

11 Final Report 11 Table 1: Implied lower risk drinking guidelines under different approaches by number of drinking days and based on mortality data Threshold Canadian: RR=1.0 RR= Relative Risk Australian: Proportion deaths attributable alcohol=1% Units per week Units per day Drinking days per week Males Females Males Females The strengths and limitations of the data on risk were considered carefully and a range of sensitivity analyses were considered by the group, in order to underpin a robust, but transparent, establishment of a low risk level for regular drinking. Table 13: Implied guideline for mean weekly consumption under different sensitivity analyses Threshold Canadian RR=1.0 RR = Relative Risk Australian Proportion =1% Drinking days per Base SA1: Threshold Males SA2: No SA3: 10 Units per week SA4: CVD Females week case effect protective yrs risks case effect protective yrs risks Base SA1: Threshold SA2: No SA3: 10 SA4: CVD Table 13, Sensitivity analyses included, based on different assumptions, are: SA1: The impact of assuming a threshold effect in the function relating peak daily consumption to risk for acute alcohol-related conditions such as injuries or accidents. The modelled threshold assumed risk equivalent to abstainers up to 3 units for females and 4 units for males. SA2: The impact of assuming no protective effects for cardiovascular conditions.

12 12 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) SA3: The impact of modelling a longer time period; namely 10 years. This allows shifts in the demographic structure of the population to occur following reduced premature mortality SA4: The impact of accounting for recent trends in mortality rates from cardiovascular conditions, using the most recent available figures from The group also considered what evidence had changed or had been added from that which underpinned the current advice that was issued in There were a number of areas that required some specific further consideration, that are discussed below, including the basis of a low risk threshold for regular drinking, the current evidence for a beneficial effect of drinking, the evidence favouring a format of weekly or daily low risk advice, and the utility of drink-free days. What evidence has changed on regular drinking guidelines? 10. The science has changed since the current guidelines came out in 1995: 11 There is much more evidence on the link between alcohol and cancer alcohol is now recognised as a cause of certain cancers by the International Agency for Research on Cancer. 12 There is an increased risk at low levels of consumption for breast, oesophageal, oral cavity, pharynx and at higher levels for e.g. liver and colorectal cancers. A consequence of this is that drinking any amount of alcohol regularly can cause harm; the risks of cancer for most people are present even at a low level of consumption, but are lower if drinking is within the proposed weekly guideline levels. Evidence for a net protective effect of alcohol from risk of death (which has been linked to possible reduced risks of heart disease late in life) is considered less strong than it was. A reduced risk still exists, but, in the UK, it now appears to matter overall in a significant way only for women aged 55 or older. 13 The 1995 report for the current guidelines found this protective effect applied at that time to men over 40 and postmenopausal women. This change in understanding is consistent with changes in the profile of heart disease in the UK and a changing population. Evidence produced concerning understanding of the core daily guideline introduced in 1995, has suggested a degree of misunderstanding by some members of the public about whether the daily limit was intended to apply as a limit for any single day or single occasion of drinking (which it was not) CMO Alcohol Guidelines Review A summary of the evidence of the health and social impacts of alcohol consumption; CMO Alcohol Guidelines Review Mapping systematic review level evidence; both reviews were written by the Centre for Public Health, Liverpool John Moores University 12 IARC Evaluations, 1988 Vol 44 (oral cavity, pharynx, larynx, oesophagus and liver); 2007 Vol 96 (colo-rectum and female breast); 2009 Vol 100 (oesophagus, head and neck) 13 Taking account of mortality from all causes. For other groups, risk reductions for death from cardiovascular conditions are either largely or wholly outweighed by risk increases for other conditions. 14 Qualitative evidence from previous Department of Health campaigns

13 Final Report 13 Recent evidence confirms that most people do not drink every day or even most days, and do not always find the current guideline, focussed on regular daily drinking, useful. 15,16 Weekly or daily guidelines? 11. We have chosen to present the analysis of the evidence in weekly terms. This is consistent with the evidence base on risk and in a simple, comprehensible form. This is in line with the principles for effective communication of health messages (to be accurate but as simple as possible). 12. There is evidence that a daily guideline may tend to be misunderstood as a maximum amount to drink on any occasion. There is evidence that only a minority of the UK population drink on more than two days each week and evidence that many find a daily guideline less useful as a result. 13. Using a weekly format also offers an inherently easier benchmark for those whose drinking amounts per day, and days per week, vary. 14. The group believes that a weekly guideline on regular drinking requires an additional recommendation, concerning the need to avoid harmful regular heavy drinking episodes, as there is clear evidence that such a pattern of drinking increases risk to health. The additional guideline we recommend is: If you do drink as much as 14 units per week, it is best to spread this evenly over 3 days or more. If you have one or two heavy drinking sessions, you increase your risks of death from long term illnesses and from accidents and injuries. The basis for determining a low risk threshold for regular drinking 15. The analysis of the evidence on levels of alcohol consumption for a number of different health harms reviewed by the group, including the evidence on cancer risks, make it apparent that there is no level of regular drinking that is completely without risks to health in the long term. 16. We have used the extensive national and international data on health risks and risks to mortality due to different levels and patterns of drinking. A key benchmark in this report is the estimate of the contribution of alcohol to risk of death. It is the point at which any apparent protective effect of drinking on mortality rate is balanced by an increased risk of death. Although there are more uncertainties concerning the level of this protective effect related to the amount consumed, the presence of the effect in the data is a consistent finding, particularly in regard to heart disease mortality, and we have taken this as a key consideration for those who wish to have a low risk of harm from drinking. 17. This is consistent with the methodology used to develop recent Canadian alcohol guidelines. Canada used a relative risk approach which set the guidelines at a point at which alcohol harms and benefits were balanced, in terms of the numbers of deaths caused and prevented. 18. The Sheffield report, which included a number of sensitivity analyses and estimates based on numbers of days drinking a week, shows that roughly similar conclusions can be reached drawing upon the Canadian relative risk approach and an absolute risk approach adapted from that used in Australia. 15 Interpretation and use of drinking guidelines and approaches by adult men and women, Presentation to the Guidelines Development Group, September 2014, University of Sheffield and Institute of Social Marketing, Stirling University 16 Lovatt M, et al. (2015) Lay epidemiology and the interpretation of low risk drinking guidelines by adults in the United Kingdom. Addiction 110

14 14 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) 19. As well as looking at the point at which the risk of death from alcohol outweighs any potential health benefit, the proposed UK guideline on regular drinking is also broadly consistent with, and takes account of, the research showing levels of drinking where alcohol would be expected to cause an overall 1% lifetime risk of death 17 for those drinking at these levels. The proposed guideline is around or a little below this level. 20. The Guidelines development group recommend that for those who drink regularly or frequently i.e. most weeks, the guideline should be not to exceed 14 units of alcohol a week. This conclusion draws on that range of analyses considered by the group described above, and the group also took account of the following considerations: the need to take account of a likely overestimation of protective effects in much of the research. mortality from ischaemic heart disease (IHD) has been falling in the UK population for many years, which means there is now less risk of death from this disease for which low alcohol consumption might give protection; this may also lead to overestimation of protective effects at the population level. the evidence for alcohol having an adverse effect on a wide range of diseases that may not cause death, which is not taken into account directly in the modelling in the Sheffield report. neither of the Canadian or Australian methods takes account of the fact that more of the deaths prevented are deaths that occur late in life (mainly heart disease), whilst many of the deaths caused by alcohol tend to occur at younger ages, for example liver disease and accidental deaths, which means that overall many more years of life are lost due to alcohol than are saved even for those drinking within the guideline level. The guideline level was determined after taking account of evidence of the risks from a range of drinking patterns including incorporating the risk to those drinking 7 days per week. Most UK adults report in surveys that they drink on 2 days a week, or less often; 18 those drinking 5 days or more a week are a small proportion of the population. Risks are higher if a given amount of alcohol is drunk on fewer days. 21. Hence, the specific selection of 14 units is based on: careful consideration of the results of the scientific modelling, the detailed analysis of risks to the UK population from drinking, consideration of what is likely to be acceptable as a low risk level of drinking, and the need to have a clear, understandable and specific message. 22. The recommendation of 14 units is both pragmatic and evidence-based. Is there a level of safe drinking? Are there acceptable risks from alcohol? 23. The 1995 guidelines have sometimes been presented as advising safe drinking levels. While they did not suggest this, the method was different in detail from either method used 17 Unlike the absolute risk method used in Australia, the method used in the Sheffield model does take account of deaths prevented, e.g. from ischaemic heart disease. So the 1% lifetime risk of death is a net risk for the population taking account of both deaths incurred and prevented. 18 Health Survey for England, 2011, Chapter 7, Drinking Diary: This is the case in England whether or not teetotallers are included in the calculation. Adult Drinking Habits in Great Britain, 2013, ONS, 2015: Those who say they drank alcohol on 5 or more days each week fell from 17% in 2005 to 11% in 2013.

15 Final Report 15 for the recent Canadian 19 and Australian 20 guidelines. It was closer to the Canadian method in looking at relative risks, but did not explicitly seek to balance harms and benefits. The 1995 report said: Most researchers have based their assessment of where the recommended upper level of alcohol consumption should be placed at the point where they judge the evidence indicates a steady increase of relative risk rising significantly from a lowest allcause mortality point on the J-shaped curve. 24. The 1995 Inter-Departmental Working Group followed a similar method. 25. A Canadian method, seeking the point where harms and benefits balance in the population as a whole, necessarily means that there is a significant body of harm, including deaths, from regular drinking within the guideline. It should be understood that at a population level the net harms and the benefits vary by age. Broadly, the net harms in terms of mortality occur under age 65 and the benefits in terms of extended life happen mostly over age The Australian method used an approach involving an assessment of absolute lifetime risk of dying from an alcohol-related condition to help them benchmark an acceptable low risk level against certain other kinds of acceptable risk-taking choices made in society (such as the risk to drivers of dying in a car accident). This is the level of drinking where alcohol would be expected to cause an overall 1% lifetime risk of death for those drinking at these levels. 27. Although we did not consider this Australian approach focussing solely on absolute risks to be adequate for our own analysis, the group did commission a report of guideline thresholds derived from the Australian approach and based on UK mortality risk estimates, using this as part of a balanced approach to determining an acceptable level of low risk, and leading to our final recommendation to drink at no more than 14 units per week would represent a low risk level. We drew upon the analyses based on both the Australian approach and the Canadian approach, alongside other analyses to test the robustness of the risk curves being used, and considered all these elements in the context of our expert knowledge of the uncertainties in the research. 28. The proposed guideline gives particular consideration to the balance of risks and benefits, but it is also set at a level around and a little below that which would be suggested by an analysis of absolute 1% lifetime risk of death. 29. Therefore, the proposed new guideline would not eliminate all health risks from drinking alcohol if all the UK population were all to drink 14 units of alcohol every week. 30. Drinking alcohol even at low levels contributes to a wide range of health harms, to a range of diseases and to hospital admissions. Hence, there is no level of drinking that can be recommended as completely safe long term, and advice to the public should be clear that many of these risks can be reduced by drinking at levels below the 14 unit weekly guideline limit, or by not drinking at all. 31. Different individuals will have different views as to what constitutes an acceptable risk of harm from their drinking. 32. We recognise that many people obtain benefits from drinking alcohol, including social pleasure. The guideline on regular drinking would be consistent with a little under a 1% lifetime risk of death from alcohol for people who follow this consistently. This implies that people need to understand that any drinking of alcohol carries some risks, like some other regular or 19 The basis for Canada s new low risk drinking guidelines: A relative risk approach to estimating hazardous levels and patterns of use, by Tim Stockwell et al. Drug and Alcohol Review (March 2012), 31, Australian Guidelines to reduce health risks from drinking alcohol, National Health and Medical Research Council, 2009

16 16 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) routine activities. 21 These are risks that people can reduce further, by choosing to drink less, or not to drink at all, if they wish. 33. This is in contrast to some other areas of safety, such as safety of drinking water, where the aim of regulation is generally to reduce health risks, such as risks of cancer, to the lowest level practicable (while avoiding disproportionate costs), as drinking water is not something people can choose to avoid. 34. Some will consider that any increased risk of death from their drinking (alongside the increased risk of illness from various alcohol-related health conditions) is worth avoiding, i.e. they would take a more cautious view than implied by these guidelines. Individual variation 35. These guidelines on regular drinking provide a guide for people in the UK as a whole, who wish to keep risks to their health from drinking at a low level. They are based on typical or average drinkers. 36. Other factors may increase the risks of harm for individuals associated with alcohol, such as: risk factors for certain diseases low body weight current health problems, or previous problems with alcohol 37. People will need to take account of these alongside the low risk weekly guidelines. 38. Similarly, for women over 55, any protective effect from death from heart disease for low levels of drinking is likely to be offset if there are other risk factors, e.g. being overweight. 39. Younger adult drinkers and younger males, in particular, have higher acute risks from drinking. 40. Middle aged or older drinkers may have accumulated risks from long term drinking, may have more relevant health conditions and in old age may be more liable to falls. 41. Differences between men and women in the immediate risks of drinking are more relevant to younger than older adults. 42. The weekly guideline is advice for the general population, which individuals need to consider in the light of their own individual characteristics as well as their own attitudes to risk. 43. It can also be seen as a benchmark to enable people to monitor their own drinking levels. 44. Figures 12 and 13 from the Sheffield report show how risks of death vary by age and frequency of drinking. 21 A close comparison is the lifetime risk of death from vehicle accidents, which was one in 240 in the UK in 2004, Department for Transport: Road Casualties Great Britain 2006 Annual Report

17 Final Report 17 Figure 12: Male relative risk of alcohol-related mortality by mean weekly consumption, number of drinking days and age 35-54: Figure 13: Female relative risk of alcohol-related mortality by mean weekly consumption, number of drinking days and age Men and women 45. It has also been possible to use available evidence to make better estimates of risks of immediate harm (accidents and injuries and deaths from these) due to drinking alcohol. These risks mainly happen with heavy drinking in a single session or in one or two days. 46. These risks are greater for men than for women, in part because of men s underlying risk taking behaviours; their risks of injury and accident are, on average, much greater than for women, even before taking account of the effects of alcohol. This means that men are more affected by increased risks of immediate harm from drinking alcohol.

18 18 Alcohol Guidelines Review Report from the Guidelines Development Group to the UK Chief Medical Officers (published January 2016 with the consultation on the language and understanding of the UK Chief Medical Officers low-risk alcohol guidelines) 47. The proposed weekly guideline is the same for men and women. Women s long-term health can be affected more by alcohol, but on average men are at much greater risk from the more immediate harms such as accidents and injuries. 48. Risks of immediate harms in women are much less than for men at the same level of consumption, but their risks of long term illness and death generally increase to a greater degree than men s as drinking increases. This is consistent with the medical understanding of general differences in biological vulnerability to exposure to alcohol over time. Reduced risks from low levels of alcohol consumption (i) Heart disease 49. The evidence 22 about any protective effect of drinking small amounts (1 unit or less a day) of alcohol in reducing risks of death, mainly from ischaemic vascular disease such as heart disease, has been taken account of in framing the regular drinking guideline and was part of the research used to inform that. 50. The expert group concluded that people who do not drink any alcohol at all should not be recommended to start drinking in the interests of their health because such advice cannot be justified for a number of reasons: (a) the evidence for a direct, protective, effect of alcohol on mortality is a subject of continuing scientific discussion; (b) methodological limitations in the evidence base mean there is uncertainty on the extent of the effect; (c) ischaemic vascular disease including heart disease, which is the key condition in the evidence of reduced risk, mainly affects older adults and particularly deaths in older age. Deaths from this type of disease have been falling in the UK population for some years, which means there is less risk for which low alcohol consumption might give protection; (d) lifestyle changes, such as stopping smoking, increasing levels of physical activity, and eating a healthy diet, can help protect against heart disease, so any potential protective effects from alcohol could be achieved in other ways, which avoid the other health risks which come with any drinking of alcohol. 51. After accounting for these limitations in the evidence used within the Sheffield model, the best specific evidence available on protective effects suggests that the maximum net reductions in deaths are present in those regularly drinking only 1 unit or less a day. 52. Previous analyses suggested the protective effect was only likely to be relevant to men from age 40 onwards and for post-menopausal women. The Sheffield report commissioned for the expert group included a UK analysis, which has found that the net protective effect that may be attributable to drinking regularly at low levels appears now to be significant only for women aged 55+ (with men aged over 55+ showing such a protective effect only of negligible size). The Sheffield report estimates that for females aged 55 and over, the greatest risk reductions occur in those drinking approximately five units per week (mean weekly consumption). 53. The impact of any such apparent protective effect would be expected to vary, for example, with differences in the risk of heart disease in the population over time, and so this recent finding is not necessarily inconsistent with previous evidence. 22 CMO Alcohol Guidelines Review A summary of the evidence of the health and social impacts of alcohol consumption; CMO Alcohol Guidelines Review Mapping systematic review level evidence; both reviews were written by the Centre for Public Health, Liverpool John Moores University